Gaming the market

I was part of a team of six who designed and implemented a system for predicting stock market prices by harvesting data from users playing a mobile app game. I wrote Go code for the server side which broadcasted financial data out to mobile app clients, and aggregated their actions within a game world to make a prediction about whether a stock price would rise or fall.

The game was a helicopter clone - the classic game of avoiding the floor and ceiling - but the floor and ceiling were being generated in real-time from financial data. The hope was that users would build up an intuition for any patterns in the underlying data.

Back to projects